- Home
- Skills
- Yuniorglez
- Gemini Elite Core
- Vercel Sync
vercel-sync_skill
- Python
7
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill yuniorglez/gemini-elite-core --skill vercel-sync- SKILL.md5.1 KB
Overview
This skill is the Vercel Sync specialist for Bun-to-Vercel orchestration, focused on Next.js 16.2 Edge-first deployments and Zero-Secret OIDC pipelines. It codifies runtime selection, build optimization, and secure CI patterns to achieve low-latency edge delivery and deterministic, zero-downtime releases.
How this skill works
The skill inspects and enforces a modern vercel.json manifest, validates bun.lock usage, and applies Edge vs Node runtime recommendations based on feature latency and resource needs. It also provides CI templates to replace long-lived tokens with OIDC, runs build and env audits, and surfaces common forensic checks when local and Vercel builds differ.
When to use it
- Deploying Next.js 16.2 apps that must run optimally on the Vercel Edge Network.
- Migrating builds to the Bun runtime and ensuring bun.lock is the single source of truth.
- Implementing secure, tokenless GitHub Actions via OIDC for Vercel deployments.
- Reducing build times using Turbopack and Bun package management.
- Deciding runtime placement for auth, streaming, and heavy native workloads.
Best practices
- Declare explicit runtime and caching directives in vercel.json and align bunVersion with CI and local environments.
- Enforce bun.lock (v2) and fail builds if package-lock.json or yarn.lock are present.
- Use OIDC in GitHub Actions (id-token: write) instead of storing VERCEL_TOKEN secrets.
- Run builds with --frozen-lockfile to ensure deterministic artifacts and avoid Vercel-local mismatches.
- Place latency-sensitive middleware on Edge, CPU-heavy streaming on Bun serverless, and native-heavy tasks on Node.js.
Example use cases
- Edge auth middleware that needs <10ms cold start and global proximity.
- AI streaming endpoints that require Bun's CPU efficiency and fast I/O for prompt streaming.
- Static pre-rendering pipelines accelerated by Bun for 2x faster build times.
- CI pipelines using OIDC to deploy previews and remove long-lived tokens from secrets stores.
- Forensic audits when a build passes locally but fails on Vercel, including bun version and vc-build log checks.
FAQ
Treat them as critical errors: remove or migrate dependencies to Bun and ensure bun.lock is committed; fail CI until resolved.
How do I eliminate VERCEL_TOKEN from GitHub Actions?
Use OpenID Connect: grant id-token: write permission, emit a short-lived token in the workflow, and call bun x vercel deploy with that token.